headroom
Warn
Audited by Socket on Oct 26, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The skill is largely coherent with its stated Headroom integration purpose, and the included code is narrowly scoped and avoids obvious exfiltration or injection. However, it installs and relies on a third-party CLI/proxy and documentation hosted on vercel.app, then modifies ~/.claude/settings.json to run a hook automatically, so the install/provenance trust is only medium and warrants caution until upstream ownership and official install documentation are independently verified.
Confidence: 85%Severity: 58%
Audit Metadata